EP3399909B1 - Method and system for determining network connections - Google Patents

Method and system for determining network connections Download PDF

Info

Publication number
EP3399909B1
EP3399909B1 EP17700316.7A EP17700316A EP3399909B1 EP 3399909 B1 EP3399909 B1 EP 3399909B1 EP 17700316 A EP17700316 A EP 17700316A EP 3399909 B1 EP3399909 B1 EP 3399909B1
Authority
EP
European Patent Office
Prior art keywords
coefficients
nodes
coherence
data set
connection
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
EP17700316.7A
Other languages
German (de)
English (en)
French (fr)
Other versions
EP3399909A1 (en
Inventor
Linda SOMMERLADE
Björn Olaf SCHELTER
Claude Michel Wischik
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Genting Taurx Diagnostic Centre Sdn Bhd
Original Assignee
Genting Taurx Diagnostic Centre Sdn Bhd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Genting Taurx Diagnostic Centre Sdn Bhd filed Critical Genting Taurx Diagnostic Centre Sdn Bhd
Priority to PL17700316T priority Critical patent/PL3399909T3/pl
Publication of EP3399909A1 publication Critical patent/EP3399909A1/en
Application granted granted Critical
Publication of EP3399909B1 publication Critical patent/EP3399909B1/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0002—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
    • A61B5/0004—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by the type of physiological signal transmitted
    • A61B5/0006—ECG or EEG signals
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0002—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
    • A61B5/0015—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system
    • A61B5/0024—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by features of the telemetry system for multiple sensor units attached to the patient, e.g. using a body or personal area network
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/369—Electroencephalography [EEG]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A61B5/7207—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal of noise induced by motion artifacts
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
    • A61B2562/04—Arrangements of multiple sensors of the same type
    • A61B2562/046—Arrangements of multiple sensors of the same type in a matrix array

Definitions

  • the present invention relates to methods and systems for determining network connections. It is particularly, but not exclusively, related to methods and systems for determining network connections in sparse networks, and has particular application to EEG data.
  • Networks of interacting nodes are a key mathematical tool for the description of complex systems (Strogatz, 2001).
  • the dynamics of the individual nodes, their coupling structure or their collective behaviour all determine the dynamics of the system.
  • detecting interactions between signals i.e. the coupling structure among nodes, is of particular interest.
  • Understanding brain networks promises to disclose the biological basis underlying natural behaviour or certain diseases (e.g. Hesse et al., 2003; Tass et al., 1998; Pitzalis et al., 1998; Keyl et al., 2000; Nollo et al., 2005; Bowers and Murray, 2004).
  • this definition states that a process x1 is causal for another process x2, if x1 is useful for the prediction of the future of x2.
  • Linear Granger-causality is typically modelled by means of vector autoregressive processes, which are estimated via multivariate Yule-Walker equations or similar approaches (Lutkepohl, 2005). In most large networks the adjacency matrix is sparse. This means that out of all the possible connections only a few are present.
  • the present inventors have realised how assuming a sparse network can be used to improve parameter estimation of the vector autoregressive process.
  • autoregressive coefficients are estimated according to Lütkepohl (2005), although other methods can be used.
  • a 15-dimensional network of coupled white noise processes is analysed.
  • Figure 1 shows the graph of the simulated network.
  • the reconstructed graphs for coherence and partial coherence are shown in Fig. 2(a) and 2(b) , respectively. Both reconstructed graphs show the same number of sub-graphs but apart from that are very different from the original one.
  • a much higher dimensional system is used. This is a 40-dimensional network of coupled white noise processes. Connections are present either at lag 1 or at lag 2.
  • the graph of the simulated network is shown in Fig. 5 .
  • N 10,000 data points of this system were simulated.
  • the estimated networks, based on coherence and partial coherence, are shown in Fig. 6(a) and (b) respectively. Neither of them yields a meaningful representation of the underlying network.
  • Electroencephalography provides multi-channel data of brain activity from a plurality (usually at least 20) small sensors attached to the scalp of an individual which detect the voltage fluctuations resulting from ionic current flows within the neurons of the brain when brain cells send messages to each other. Due to the plurality of sensors, the data from EEG has a corresponding plurality of channels. EEG is currently used to help diagnose and monitor a number of conditions affecting the brain, particularly epilepsy.
  • EEG EEG has been employed clinically as a measure of brain function in the hope of determining and differentiating certain functional conditions of the brain.
  • progress has been slow.
  • the present inventors have realised that the possibility of accurately determining networks of causal relationships within EEG data may allow further interpretation of this data for clinical purposes.
  • EEG data is multi-dimensional and so susceptible to analysis in a network fashion. Many types of EEG data now have a large number of channels (e.g. 20 or more). Traditional methods of analysis can struggle to provide meaningful information or interpretation of such multi-variate data and so new approaches are required to handle the multi-order systems which are being observed.
  • Alzheimer's disease is a long time (possibly of the order of 20 years) in development before frank symptoms are observed. Accordingly, there is considerable focus on techniques which may provide for reliable early-stage identification of potential suffers or of those exhibiting particular susceptibility or risk factors. These techniques aim to identify the onset of early-stage Alzheimer's disease in the entirely pre-clinical stage (typically 10-20 years before onset) or in the prodromal period.
  • EEG data to identify early symptoms or warning signs of Alzheimer's disease has particular attractions if suitable methods exist for the robust analysis of the data as it is a standard and widely-used and available technique. It is also available, in certain implementations, in a form which can be effectively self-applied by an individual with little or no training and so is eminently suitable for the primary care setting.
  • the present invention aims to provide methods and systems which provide accurate and reliable predictions of network connections and coefficients, particularly in sparse networks.
  • a further aim of the present invention is to provide efficient methods for predicting network connections and coefficients.
  • US 2013/0123607 A1 proposes a computing device for use in a system for mapping brain activity of a subject includes a processor.
  • the processor is programmed to select a plurality of measurements of brain activity that is representative of at least one parameter of a brain of the subject during a resting state.
  • the processor is programmed to compare at least one data point from each of the measurements with a corresponding data point from a previously acquired data set from at least one other subject.
  • the processor is also programmed to produce at least one map for each of the measurements based on the comparison of the resting state data point and the corresponding previously acquired data point.
  • the processor may also be programmed to categorize the brain activity in a plurality of networks in the brain based on the map.
  • Multivariate partial coherence analysis for identification of neuronal connectivity from multiple electrode array recordings proposes a method to infer the connectivity between spiking neurons.
  • a network of unconditional relationships between ten neuronal spikes and a network of conditional relationships between the same spikes are obtained by calculating coherence and partial coherence values between the spikes, respectively, as weighted links in the first and second networks.
  • Respective significance threshold values are then applied to the link weightages in each network.
  • a further aim of the present invention is to provide methods and systems for processing EEG data to provide meaningful information and/or interpretation of the data through network overviews of brain activity and to permit subsequent uses of such data.
  • aspects of the present invention provide methods and systems for identifying connections between nodes in a network which operate by identifying probable zero connection coefficients and setting them to zero for subsequent processing.
  • a first aspect of the present invention provides a computer-implemented method of monitoring brain function in a patient, the method including identifying the network connections between the node signals on an EEG recording performed on the patient over a period of time, by using a method of identifying, in a network of interacting nodes simultaneously producing signals, connections between said nodes and of estimating connection coefficients between nodes identified as connected, the method including the steps of: periodically recording the signal at each node over a predetermined period of time to form a data set; calculating the coherence and partial coherence between each combination of nodes in the data set; checking, for each combination of nodes, if either the coherence or the partial coherence is below a first predetermined threshold and, if so, setting the corresponding connection coefficient to zero for all subsequent steps; a first estimating step of estimating, from the data set, the connection coefficients for the combinations of nodes for which the connection coefficient has not already been set to zero; for each connection coefficient estimated by said first estimating step to be below a second threshold, setting
  • the method of this aspect contains two "zeroing" steps prior to the final step of estimating the connection coefficients. These act to remove indirect links (by consideration of partial coherence and coherence). This can reduce or eliminate false positives from the determined network.
  • the method of this aspect just uses data. It does not rely on any predictions or assumptions about the underlying model.
  • the method of this aspect assumes a degree of sparseness in the network, implying that some of the coefficients in the adjacency matrix are zero. Based on this assumption the estimation procedure is improved by identifying coefficients which are candidates for zeroing and setting these to zero before performing further calculations or estimations.
  • the method of this aspect is therefore also more efficient at predicting connections in a sparse network than existing methods. Accordingly the method is preferably applied to networks which are known or predicted to be sparsely connected.
  • sparsely connected we mean that the network has at least 50% of potential connections between pairs of nodes which are not present (i.e. the connection coefficients are zero), preferably at least 60% and in some embodiments at least 75%. Indeed, the method of this aspect becomes more efficient as the network becomes sparser, and so can be applied to networks in which 80% or 90% of the potential connections are not present.
  • This method enables application of Granger-causality in high-dimensional systems (in particular those having 10 or more nodes).
  • existing Granger-causality inferences typically work well in low-dimensional systems
  • the additional steps of the present method in reducing the number of relevant coefficients allows the application of Granger-causality in higher-dimensional systems, particularly (but not exclusively) where these systems are sparsely connected.
  • the method of this aspect can enable a reliable estimation of Granger-causality.
  • the method of this aspect can be readily applied to various measures for Granger causality and other approaches that are based on vector autoregressive models.
  • the step of checking involves calculating the product of the calculated coherence and partial coherence for each combination of nodes and determining whether said product is below said first predetermined threshold. If either of the coherence or the partial coherence is zero, or close to zero, then the resulting product will be zero, or close to zero. This means that, for each coefficient, only a single comparison with the threshold is required.
  • the first predetermined threshold may be the critical value for partial coherence as defined in Schad et al (2009). Alternatively, the first predetermined threshold may be the critical value for coherence. Alternatively, the first predetermined threshold may be the product of both the critical value for partial coherence and the critical value for coherence.
  • the second threshold for the estimated connection coefficients may be determined by: separating the estimated coefficients into two groups according to the squared Euclidean distance of the estimated coefficients, a first group containing those coefficients with high values of the coefficients and the second group containing those coefficients with low values; and setting said second threshold as a value which is greater than the value of all coefficients in said second group.
  • the estimated connection coefficients can be separated into two groups and the separation between weak connections and stronger connections, thus permitting a clear distinction to be drawn between the estimated connections which are likely to be entirely due to noise and those which represent a genuine connection.
  • the second threshold may therefore be variably chosen at the appropriate level to separate these two groups. Alternatively, the second threshold may be set in advance.
  • the first and second estimating steps for estimating the connection coefficients estimate the autoregressive coefficients of the data set.
  • the method further includes the step of screening the data set to remove outliers.
  • the method of this aspect works on measured data, it may be susceptible to outliers in the data.
  • Outliers are artefacts in the data generally caused by events which are not intended to be measured as part of the recorded data. For example, in EEG data, eye-blinks can cause artefacts of this kind. Removing such outliers from the data set can therefore improve the accuracy of the method.
  • the method further includes the step of filtering the data set to remove noise.
  • filtering the data set to remove noise can improve the accuracy of the method.
  • the screening or filtering can be performed prior to the step of calculating, or could be incorporated into the actual estimation of the coefficients.
  • the network of interacting nodes producing signals are an electroencephalographic (EEG) system.
  • EEG electroencephalographic
  • Network structure analysis on EEG data can provide an insight into both brain activity and muscle activity and the resulting networks can be used for comparative purposes, for example against sample networks for particular populations, or as comparators for future studies on the same individual.
  • the temporal resolution of EEG is in the millisecond range. It is known that brain processing time is of the order of 500ms and so the method of the present is preferably applied to this data. However, the techniques are equally applicable to other data with lower resolution (e.g. functional magnetic resonance imaging or fRMI which has a temporal resolution of approximately 2s).
  • EEG data is currently typically recorded over 20 minute periods. This can lead to practical data collection issues in observing the patient in a constant state (or plurality of states) over such a time period, as well as increasing the probability of artefacts arising. If the time period can be reduced further, perhaps to a few 100 seconds, then these problems can be reduced and/or avoided.
  • the method of the present aspect can provide a robust prediction of the network from relatively small quantities of data, the amount of EEG data (and therefore the length of time) needed can potentially be reduced.
  • the method of the present aspect may include any combination of some, all or none of the above described preferred and optional features.
  • a second aspect of the present invention provides a method of monitoring brain function in a patient, the method including the steps of: performing an EEG recording on a patient over a period of time; identifying the network connections between the node signals on the EEG using a method according to the above described first aspect, including some, all or none of the optional or preferred features of that aspect.
  • a third aspect of the present invention provides a system for identifying network connections and estimating connection coefficients between nodes in a data recording of brain activity, the system comprising: a plurality of sensors for recording the brain activity of an individual at different locations over a predetermined period of time to produce a data set; and a processor which is configured to: calculate the coherence and partial coherence between each combination of nodes in the data set; check, for each combination of nodes, if either the coherence or the partial coherence is below a first predetermined threshold and, if so, set the corresponding connection coefficient to zero for all subsequent steps; estimate, from the data set, the connection coefficients for the combinations of nodes for which the connection coefficient has not already been set to zero; for each connection coefficient estimated to be below a second threshold, set said coefficient to zero for all subsequent steps; and re-estimate, from the data set, the connection coefficients for the combinations of nodes for which the connection coefficients have not already been set to zero.
  • the system of this aspect processes the recorded data and applies two "zeroing" steps prior to the final step of estimating the connection coefficients. These act to remove indirect links (by consideration of partial coherence and coherence). This can reduce or eliminate false positives from the determined network.
  • the system of this aspect just uses data. It does not rely on any predictions or assumptions about the underlying model.
  • the system of this aspect assumes a degree of sparseness in the network, implying that some of the coefficients in the adjacency matrix are zero. Based on this assumption the estimation procedure is improved by identifying coefficients which are candidates for zeroing and setting these to zero before performing further calculations or estimations.
  • the method of this aspect is therefore also more efficient at predicting connections in a sparse network than existing methods. Accordingly the method is preferably applied to networks which are known or predicted to be sparsely connected.
  • sparsely connected we mean that the network has at least 50% of potential connections between pairs of nodes which are not present (i.e. the connection coefficients are zero), preferably at least 60% and in some embodiments at least 75%. Indeed, the system of this aspect becomes more efficient as the network becomes sparser, and so can be applied to networks in which 80% or 90% of the potential connections are not present.
  • This processor of this system applies Granger-causality in the high-dimensional system (in particular one having 10 or more nodes). Although existing Granger-causality of data could work well in low-dimensional systems, the processing to reduce the number of relevant coefficients allows the application of Granger-causality in such higher-dimensional systems, particularly (but not exclusively) where these systems are sparsely connected.
  • the system of this aspect can reliably estimate Granger-causality and can be readily applied to various measures for Granger causality but also to other approaches that are based on vector autoregressive models.
  • Network structure analysis on brain activity data can provide an insight into both brain activity and muscle activity and the resulting networks can be used for comparative purposes, for example against sample networks for particular populations, or as comparators for future studies on the same individual.
  • the processor is configured to calculate the product of the calculated coherence and partial coherence for each combination of nodes and determine whether said product is below said first predetermined threshold. If either of the coherence or the partial coherence is zero, or close to zero, then the resulting product will be zero, or close to zero. This means that, for each coefficient, only a single comparison with the threshold is required.
  • the first predetermined threshold may be the critical value for partial coherence as defined in Schad et al (2009). Alternatively, the first predetermined threshold may be the critical value for coherence. Alternatively, the first predetermined threshold may be the product of both the critical value for partial coherence and the critical value for coherence.
  • the processor may be configured to determine said second threshold by: separating the estimated coefficients into two groups according to the squared Euclidean distance of the estimated coefficients, a first group containing those coefficients with high values and the second group containing those coefficients with low values; and setting said second threshold as a value which is greater than the value of all coefficients in said second group.
  • the estimated connection coefficients can be separated into two groups and the separation between weak connections and stronger connections, thus permitting a clear distinction to be drawn between the estimated connections which are likely to be entirely due to noise and those which represent a genuine connection.
  • the second threshold may therefore be variably chosen at the appropriate level to separate these two groups. Alternatively, the second threshold may be set in advance.
  • the processor is configured to estimate the connection coefficients by estimating the autoregressive coefficients of the data set.
  • the processor is configured to screen the data set to remove outliers.
  • the processor of this aspect is processing measured brain activity data, it may be susceptible to outliers in the data.
  • Outliers are artefacts in the data generally caused by events which are not intended to be measured as part of the recorded data. For example, in EEG data, eye-blinks in particular can cause artefacts of this kind. Removing such outliers from the data set can therefore improve the accuracy of the system.
  • the processor is configured to filter the data set to remove noise.
  • the processor of this aspect is processing measured brain activity data, it may be susceptible to noise in those measurements. Accordingly, filtering the data set to remove noise can improve the accuracy of the system.
  • the screening or filtering can be performed prior to the step of calculating, or could be incorporated into the actual estimation of the coefficients.
  • the system is applied to electroencephalographic (EEG) data and the plurality of sensors are an electroencephalograph.
  • EEG electroencephalographic
  • the temporal resolution of EEG is in the millisecond range. It is known that brain processing time is of the order of 500ms and so the method of the present is preferably applied to this data. However, the techniques are equally applicable to other data with lower resolution (e.g. functional magnetic resonance imaging or fRMI which has a temporal resolution of approximately 2s).
  • EEG data is currently typically recorded over 20 minute periods. This can lead to practical data collection issues in observing the patient in a constant state (or plurality of states) over such a time period, as well as increasing the probability of artefacts arising. If the time period can be reduced further, perhaps to a few 100 seconds, then these problems can be reduced and/or avoided.
  • the system of the present aspect can provide a robust prediction of the network from relatively small quantities of data, the amount of EEG data (and therefore the length of time) needed can potentially be reduced.
  • the system of the present aspect may include any combination of some, all or none of the above described preferred and optional features.
  • the system of the present aspect may operate by carrying out a method according to the above first or second aspects of this invention, but need not do so.
  • a method according to an embodiment of the present invention is shown schematically in the flow chart of Fig. 8 .
  • the method involves a three step approach to estimate sparse autoregressive processes.
  • the underlying principle is to rule out some of the coefficients before the actual fitting procedure.
  • x ⁇ pCS xy
  • the threshold which is used to determine whether a coefficient is compatible with zero may be the critical value for partial coherence or the critical value for coherence, as defined in Schad et al. (2009).
  • the resulting coefficients are separated into two clusters according to their squared euclidean distance.
  • the coefficients in the cluster with the smaller values are then set to zero (S105). This step accounts for the fact that coherence and partial coherence are symmetric measures and can therefore not rule out single directed connections.
  • autoregressive coefficients are estimated again (Eqs. (5) and (6) - S106). This time all coefficients identified to be compatible with zero in either the first or second step are kept at zero.
  • the key benefit of this method is that only the non-zero coefficients are estimated in the third step. Since the number of estimated coefficients is dramatically reduced by this procedure, the accuracy of the estimation can be improved. Meaning that less coefficients are estimated from the same number of data points.
  • Fig. 11 is the equivalent to Fig. 4 and shows, for all 36 coefficients, the absolute value of the difference between the true and the mean estimated coefficient using the method according to the above embodiment.
  • Error bars refer to the standard deviation of the mean for 100 realisations.
  • Coefficients that are displayed without error bar are those that were fixed at zero, all of which are truly zero in the simulation. The result shows that the six non-zero coefficients are estimated very close to their respective true value (and closer than in the previously-described process), whilst the 30 coefficients that are zero are all correctly identified as exactly zero.
  • Fig. 12 shows a further simulated 15-dimensional network of coupled white noise processes.
  • Fig. 13 shows the connections and coefficients of the network of Fig. 12 as predicted using coherence alone. It can be seen that, as well as adding a number of additional connections within the more complex part of the network, this approach also mis-characterises the relationship between nodes 1, 6, 8 and 10.
  • Fig. 14 shows the connections and coefficients of the network of Fig. 12 as predicted using partial-coherence. This approach is more successful as it removes a number of indirect links. However, it still incorrectly predicts connections in the more complex part of the network which are not present in the underlying network (false positives) and, like the coherence approach, mis-characterises the relationship between nodes 1, 6, 8 and 10.
  • Fig. 15 shows the connections and coefficients of the network of Fig. 12 as calculated using the method of the embodiment described above. It can be seen from a comparison with Fig. 12 that the network is completely accurately mapped.
  • Fig. 16 shows a further simulated 40-dimensional network of coupled white noise processes.
  • Fig. 17 shows the predicted network using coherence, and it can be seen that as the dimension of the network increases, the number of false positives generated in this approach can completely overwhelm the true connections.
  • Fig. 18 shows the predicted network using partial coherence. As with the previous simulation, this cuts out a significant number of false positive connections, but still does not completely reproduce the underlying network. The method of the embodiment described above accurately reproduces the network shown in Fig. 16 .
  • EEG electroencephalogram
  • Figs. 20 and 21 illustrate, schematically, the kinds of networks that would be expected for a healthy individual and an individual with mild cognitive impairment.
  • the nodes have been simplified to the frontal, central, temporary and occipital regions (4 channel EEG).
  • a healthy individual would be expected to have interconnections between all of the regions and within each individual region (as the nodes are effectively identical, they have not been labelled in Fig. 20 ).
  • Fig. 21 shows that, in an individual with mild cognitive impairment, whilst there is connection within each region, the interconnection between each region is significantly reduced, and generally only involves flow in one direction from frontal to occipital with little or no return flow from the occipital or temporal regions, and no connections between non-adjacent regions.
  • the sparse estimation technique of the present invention can readily be applied to frequency-domain measures for Granger-causality such as re-normalised partial directed coherence (Schelter et al., 2009). This will allow investigating networks at different frequencies which is of particular interest in EEG analysis.
  • the method of the embodiments set out above is data driven. This means that the sparse coefficient matrix is estimated based on a given data set measured from the nodes. Zeros are placed according to coherence and partial coherence estimated from the data set. Should the underlying network not be sparse, the algorithm will simply proceed to estimate all coefficients in the way that the underlying estimation procedure does.
  • measured data can be afflicted by outliers.
  • these outliers include eye-blink artefacts.
  • outliers such as eye-blinks can be removed using weighted robust Kalman filtering (Ting et al., 2007) or the more general outlier robust Kalman filtering (Agamennoni et al., 2011). The methods described herein can be combined with these approaches to further improve the estimation technique.
  • Methods of predicting according to embodiments of this invention can be used in a variety of settings. As shown above, they have particular use in the analysis of EEG data and, from the predicted networks, it is possible to carry out further assessments or make determinations of the cognitive function of the individual from whom the data is taken.
  • the network prediction methods can be used in a variety of ways.
  • diagnosis for example by comparison of the network obtained from EEG data from an individual against comparative networks for healthy individuals and those with cognitive impairments
  • the network predictions can be used to monitor the response of an individual to treatment for cognitive impairment.
  • a record of the networks derived from EEG data from an individual undergoing treatment can be maintained and changes monitored over time.
  • An effective treatment may either slow or stop the deterioration in the individual's cognitive function (as represented by the number and/or strength of connections in the network), or may result in a reversal of previous decline (as represented by an increase in the number and/or strength of connections in the network).
  • the network prediction methods can also be used as a discriminator to identify or screen candidates for trials of a particular treatment by allowing identification of individuals with particular characteristics of cognitive function at the neural level.
  • the network prediction methods can also be used to screen to ensure either that all participants in the trial have the same, or similar, cognitive impairments, or that participants having a wide range of cognitive impairments are selected.
  • the network prediction methods can also be used in conjunction with targeted response questions to test specific responses and determine whether this affects the pattern of responses within the brain during that testing.
  • a computer system includes the hardware, software and data storage devices for embodying a system or carrying out a method according to the above described embodiments.
  • a computer system may comprise a central processing unit (CPU), input means, output means and data storage.
  • the computer system has a monitor to provide a visual output display.
  • the data storage may comprise RAM, disk drives or other computer readable media.
  • the computer system may include a plurality of computing devices connected by a network and able to communicate with each other over that network.
  • the methods of the above embodiments may be provided as computer programs or as computer program products or computer readable media carrying a computer program which is arranged, when run on a computer, to perform the method(s) described above.
  • computer readable media includes, without limitation, any non-transitory medium or media which can be read and accessed directly by a computer or computer system.
  • the media can include, but are not limited to, magnetic storage media such as floppy discs, hard disc storage media and magnetic tape; optical storage media such as optical discs or CD-ROMs; electrical storage media such as memory, including RAM, ROM and flash memory; and hybrids and combinations of the above such as magnetic/optical storage media.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Pathology (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Surgery (AREA)
  • Veterinary Medicine (AREA)
  • Biophysics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Molecular Biology (AREA)
  • Psychiatry (AREA)
  • Signal Processing (AREA)
  • Data Mining & Analysis (AREA)
  • Physiology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Psychology (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Theoretical Computer Science (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Algebra (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Software Systems (AREA)
  • Power Engineering (AREA)
EP17700316.7A 2016-01-08 2017-01-06 Method and system for determining network connections Active EP3399909B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PL17700316T PL3399909T3 (pl) 2016-01-08 2017-01-06 Sposób i system do określania połączeń sieci

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
MYPI2016000032 2016-01-08
PCT/EP2017/050269 WO2017118733A1 (en) 2016-01-08 2017-01-06 Method and system for determining network connections

Publications (2)

Publication Number Publication Date
EP3399909A1 EP3399909A1 (en) 2018-11-14
EP3399909B1 true EP3399909B1 (en) 2022-03-02

Family

ID=57796338

Family Applications (1)

Application Number Title Priority Date Filing Date
EP17700316.7A Active EP3399909B1 (en) 2016-01-08 2017-01-06 Method and system for determining network connections

Country Status (16)

Country Link
US (1) US11006830B2 (da)
EP (1) EP3399909B1 (da)
JP (1) JP6926117B2 (da)
KR (1) KR20180100154A (da)
CN (1) CN108463165B (da)
AU (1) AU2017205100B2 (da)
CA (1) CA3009874A1 (da)
DK (1) DK3399909T3 (da)
ES (1) ES2909135T3 (da)
HK (1) HK1258178A1 (da)
PL (1) PL3399909T3 (da)
PT (1) PT3399909T (da)
RU (1) RU2732435C2 (da)
SG (1) SG11201805352QA (da)
TW (1) TWI745321B (da)
WO (1) WO2017118733A1 (da)

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11615285B2 (en) 2017-01-06 2023-03-28 Ecole Polytechnique Federale De Lausanne (Epfl) Generating and identifying functional subnetworks within structural networks
CN108523907B (zh) * 2018-01-22 2021-07-16 上海交通大学 基于深度收缩稀疏自编码网络的疲劳状态识别方法及系统
US11663478B2 (en) 2018-06-11 2023-05-30 Inait Sa Characterizing activity in a recurrent artificial neural network
US11972343B2 (en) 2018-06-11 2024-04-30 Inait Sa Encoding and decoding information
US11893471B2 (en) 2018-06-11 2024-02-06 Inait Sa Encoding and decoding information and artificial neural networks
WO2020048593A1 (en) 2018-09-05 2020-03-12 Wista Laboratories Ltd. Network methods for neurodegenerative diseases
US11836215B2 (en) * 2019-03-07 2023-12-05 Axel W. E. Wismüller Method and device for determining a measure of causal influence between components of complex systems
US11569978B2 (en) 2019-03-18 2023-01-31 Inait Sa Encrypting and decrypting information
US11652603B2 (en) 2019-03-18 2023-05-16 Inait Sa Homomorphic encryption
WO2020207983A1 (en) * 2019-04-10 2020-10-15 Genting Taurx Diagnostic Centre Sdn Bhd Adaptive neurological testing method
US10706104B1 (en) * 2019-07-25 2020-07-07 Babylon Partners Limited System and method for generating a graphical model
US11816553B2 (en) 2019-12-11 2023-11-14 Inait Sa Output from a recurrent neural network
US11651210B2 (en) * 2019-12-11 2023-05-16 Inait Sa Interpreting and improving the processing results of recurrent neural networks
US11580401B2 (en) 2019-12-11 2023-02-14 Inait Sa Distance metrics and clustering in recurrent neural networks
US11797827B2 (en) 2019-12-11 2023-10-24 Inait Sa Input into a neural network
CN112244870B (zh) * 2020-09-24 2022-02-22 杭州电子科技大学 基于符号化排列传递熵的癫痫脑电双向耦合分析方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7415305B2 (en) * 2004-10-01 2008-08-19 The Trustees Of Columbia University In The City Of New York Method for the spatial mapping of functional brain electrical activity
US9480402B2 (en) * 2011-11-11 2016-11-01 Washington University System and method for task-less mapping of brain activity
US8977029B2 (en) * 2012-08-24 2015-03-10 Siemens Aktiengesellschaft Method and system for multi-atlas segmentation of brain computed tomography image data
TWI503103B (zh) * 2013-04-12 2015-10-11 Inst Nuclear Energy Res A method of stimulating the formation of brain cognitive response images
TW201521676A (zh) * 2013-12-13 2015-06-16 Nat Inst Chung Shan Science & Technology 一種使用類神經網路產生判斷麻醉意識清醒程度指標的方法
US9107595B1 (en) * 2014-09-29 2015-08-18 The United States Of America As Represented By The Secretary Of The Army Node excitation driving function measures for cerebral cortex network analysis of electroencephalograms
CN104715150A (zh) * 2015-03-19 2015-06-17 上海海事大学 一种基于复杂网络的偏头痛患者大脑皮层的辅助分类分析方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MAKHTAR SITI N ET AL: "Multivariate partial coherence analysis for identification of neuronal connectivity from multiple electrode array recordings", 2014 IEEE CONFERENCE ON BIOMEDICAL ENGINEERING AND SCIENCES (IECBES), IEEE, 8 December 2014 (2014-12-08), pages 77 - 82, XP032739016, DOI: 10.1109/IECBES.2014.7047613 *

Also Published As

Publication number Publication date
JP6926117B2 (ja) 2021-08-25
AU2017205100A1 (en) 2018-07-26
CN108463165B (zh) 2021-08-24
RU2018128580A (ru) 2020-02-10
AU2017205100B2 (en) 2021-05-27
RU2018128580A3 (da) 2020-04-09
KR20180100154A (ko) 2018-09-07
TW201725519A (zh) 2017-07-16
TWI745321B (zh) 2021-11-11
US20190021594A1 (en) 2019-01-24
EP3399909A1 (en) 2018-11-14
US11006830B2 (en) 2021-05-18
JP2019507450A (ja) 2019-03-14
PL3399909T3 (pl) 2022-06-20
SG11201805352QA (en) 2018-07-30
HK1258178A1 (zh) 2019-11-08
DK3399909T3 (da) 2022-03-21
CA3009874A1 (en) 2017-07-13
WO2017118733A1 (en) 2017-07-13
RU2732435C2 (ru) 2020-09-16
CN108463165A (zh) 2018-08-28
ES2909135T3 (es) 2022-05-05
PT3399909T (pt) 2022-03-24

Similar Documents

Publication Publication Date Title
EP3399909B1 (en) Method and system for determining network connections
Fan et al. Detecting abnormal pattern of epileptic seizures via temporal synchronization of EEG signals
Bhardwaj et al. A novel genetic programming approach for epileptic seizure detection
CA2875912C (en) Assessing susceptibility to epilepsy and epileptic seizures
US20130178731A1 (en) Methods and apparatus for risk assessment of developmental disorders during early cognitive development
Rana et al. Seizure detection using the phase-slope index and multichannel ECoG
Song et al. A new neural mass model driven method and its application in early epileptic seizure detection
WO2016113718A1 (en) System and method to diagnose and predict different systemic disorders and mental states
US20060287607A1 (en) Multi-dimensional dynamical analysis
Vandana et al. A review of EEG signal analysis for diagnosis of neurological disorders using machine learning
Alzahy et al. Optimal EEG window size for neural seizure detection
Ansari et al. A review on automated detection of non-convulsive seizures using eeg
Gu et al. Detecting epileptic seizures via non-uniform multivariate embedding of EEG signals
Pampu et al. Transfer entropy as a tool for reconstructing interaction delays in neural signals
SOMMERLADE et al. Patent 3009874 Summary
Bates et al. Seizure detection by recurrent backpropagation neural network analysis
Bakhtyari et al. Detection of ADHD Disorder Using Dynamic Connectivity Tensors in Bidirectional Circular Reservoir Computing
Rajabioun Autistic recognition from EEG signals by extracted features from several time series models
Alves et al. Microstate Graphs: A Node-Link Approach to Identify Patients with Schizophrenia
Nattudurai Classifying Schizophrenia Disorder through EEG Signal Analysis and Machine Learning
Sclabassi Robyn R. Bates, Mingui Sun, Mark L. Scheuer
Bates et al. Analysis of multi-channel subdural EEG by recurrent neural networks
MODEL et al. Alma Mater Studiorum· Universita di Bologna
Jinxu et al. Functional EEG Network Analysis for Cognitive Diagnosis of Alzheimer’s Disease
Bates et al. Neural network analysis of seizure activity in subdural EEG

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20180620

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

AX Request for extension of the european patent

Extension state: BA ME

DAV Request for validation of the european patent (deleted)
DAX Request for extension of the european patent (deleted)
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: EXAMINATION IS IN PROGRESS

17Q First examination report despatched

Effective date: 20190524

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: EXAMINATION IS IN PROGRESS

REG Reference to a national code

Ref country code: HK

Ref legal event code: DE

Ref document number: 1258178

Country of ref document: HK

RIC1 Information provided on ipc code assigned before grant

Ipc: A61B 5/369 20210101AFI20210222BHEP

Ipc: A61B 5/00 20060101ALN20210222BHEP

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: GRANT OF PATENT IS INTENDED

INTG Intention to grant announced

Effective date: 20210412

GRAJ Information related to disapproval of communication of intention to grant by the applicant or resumption of examination proceedings by the epo deleted

Free format text: ORIGINAL CODE: EPIDOSDIGR1

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: EXAMINATION IS IN PROGRESS

REG Reference to a national code

Ref country code: DE

Ref legal event code: R079

Ref document number: 602017053982

Country of ref document: DE

Free format text: PREVIOUS MAIN CLASS: A61B0005047600

Ipc: A61B0005369000

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: GRANT OF PATENT IS INTENDED

INTC Intention to grant announced (deleted)
RIC1 Information provided on ipc code assigned before grant

Ipc: A61B 5/00 20060101ALN20210816BHEP

Ipc: A61B 5/369 20210101AFI20210816BHEP

INTG Intention to grant announced

Effective date: 20210906

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE PATENT HAS BEEN GRANTED

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

Ref country code: AT

Ref legal event code: REF

Ref document number: 1471596

Country of ref document: AT

Kind code of ref document: T

Effective date: 20220315

REG Reference to a national code

Ref country code: DE

Ref legal event code: R096

Ref document number: 602017053982

Country of ref document: DE

REG Reference to a national code

Ref country code: DK

Ref legal event code: T3

Effective date: 20220318

REG Reference to a national code

Ref country code: PT

Ref legal event code: SC4A

Ref document number: 3399909

Country of ref document: PT

Date of ref document: 20220324

Kind code of ref document: T

Free format text: AVAILABILITY OF NATIONAL TRANSLATION

Effective date: 20220317

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: SE

Ref legal event code: TRGR

REG Reference to a national code

Ref country code: ES

Ref legal event code: FG2A

Ref document number: 2909135

Country of ref document: ES

Kind code of ref document: T3

Effective date: 20220505

REG Reference to a national code

Ref country code: NL

Ref legal event code: FP

REG Reference to a national code

Ref country code: LT

Ref legal event code: MG9D

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: RS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: NO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220602

Ref country code: LT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: HR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220602

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LV

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220603

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SM

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220702

Ref country code: AL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

REG Reference to a national code

Ref country code: DE

Ref legal event code: R097

Ref document number: 602017053982

Country of ref document: DE

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20221216

Year of fee payment: 7

26N No opposition filed

Effective date: 20221205

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20220302

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: IT

Payment date: 20230125

Year of fee payment: 7

Ref country code: BE

Payment date: 20221228

Year of fee payment: 7

REG Reference to a national code

Ref country code: AT

Ref legal event code: UEP

Ref document number: 1471596

Country of ref document: AT

Kind code of ref document: T

Effective date: 20220302

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LU

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20230106

P01 Opt-out of the competence of the unified patent court (upc) registered

Effective date: 20231025

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: SE

Payment date: 20231220

Year of fee payment: 8

Ref country code: PT

Payment date: 20231011

Year of fee payment: 8

Ref country code: NL

Payment date: 20231220

Year of fee payment: 8

Ref country code: DK

Payment date: 20231220

Year of fee payment: 8

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: PL

Payment date: 20231011

Year of fee payment: 8

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: IE

Payment date: 20240122

Year of fee payment: 8

Ref country code: ES

Payment date: 20240201

Year of fee payment: 8

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: AT

Payment date: 20231222

Year of fee payment: 8

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20240122

Year of fee payment: 8

Ref country code: GB

Payment date: 20240122

Year of fee payment: 8

Ref country code: CH

Payment date: 20240202

Year of fee payment: 8